In message , t2t writes
stop farting ****e then at Force 5.
didnt u hear the shipping forecast



Does your mother, or nurse, know what you're up to on this here
interwebby thing?

You naughty boy, back to special school soon, never mind eh?
--


Jim

"Dawlish" wrote in message
...
On Aug 27, 11:47 am, "David Haggas"
wrote:

FWIW, I noticed no-one replied to your request for wind strengths. At
that distance, there are far more skilled people than me at prediting
that; I wouldn't dream of competing with the MetO at any closer than
10 days, but I fear it was just too far out to predict with accuracy
when you asked. Maybe ask again.

It was another poster that asked the question about wind. I tend to use the
excellent XC weather site (GFS).

"Jim Kewley" wrote in message
news
In message , t2t writes
stop farting ****e then at Force 5.
didnt u hear the shipping forecast



Does your mother, or nurse, know what you're up to on this here interwebby
thing?

You naughty boy, back to special school soon, never mind eh?
--


Jim

Jim where have you been? Did Kenneth Clarke finally respond to those
letters?

On Aug 28, 9:10*am, "David Haggas" wrote:
"Dawlish" wrote in message

...
On Aug 27, 11:47 am, "David Haggas"
wrote:

FWIW, I noticed no-one replied to your request for wind strengths. At
that distance, there are far more skilled people than me at prediting
that; I wouldn't dream of competing with the MetO at any closer than
10 days, but I fear it was just too far out to predict with accuracy
when you asked. Maybe ask again.

It was another poster that asked the question about wind. I tend to use the
excellent XC weather site (GFS).

Apologies, David, I lost it in the postings.

If XC weather do manage to forecast accurate wind speeds, for
particular localities 5 days in advance, they are doing incredibly
well. The fact that they forecast them - and it is a very well set out
site and interesting - does not, of course, mean the forecasting is
accurate. That can only be determined by long-term output percentage
accuracy stats.

On Aug 28, 12:38*pm, Dawlish wrote:
On Aug 28, 9:10*am, "David Haggas" wrote:

"Dawlish" wrote in message

...
On Aug 27, 11:47 am, "David Haggas"
wrote:

FWIW, I noticed no-one replied to your request for wind strengths. At
that distance, there are far more skilled people than me at prediting
that; I wouldn't dream of competing with the MetO at any closer than
10 days, but I fear it was just too far out to predict with accuracy
when you asked. Maybe ask again.

It was another poster that asked the question about wind. I tend to use the
excellent XC weather site (GFS).

Apologies, David, I lost it in the postings.

If XC weather do manage to forecast accurate wind speeds, for
particular localities 5 days in advance, they are doing incredibly
well. The fact that they forecast them - and it is a very well set out
site and interesting - does not, of course, mean the forecasting is
accurate. That can only be determined by long-term output percentage
accuracy stats.

More indications that the re-intrusion of a zonal Atlantic influence
may be put back further, on the models tonight.

On Aug 24, 10:18*pm, Dawlish wrote:
Well! I can't ignore that, as my forecast criteria have been triggered
by cross-model agreement and consistency from both the gfs and the ECM
today:

** At T240, on 3rd Sept, the UK will be bathed in a warm airstream,
with a continental feed. Higher pressure will be over, or to the east
of the UK and lower pressure will be kept out in the Atlantic, or
towards Iceland, by the bocking high. This situation will have been
preceded by a *period of dry and settled weather, starting towards the
end of the last week in August and including the Bank Holiday. **

Short and simple. I don't see an early autumn at all. The unsettled
weather will continue this week, but by early September, there will be
(questionable, for sure!) talk of an "Indian summer" (whatever that
actually is!) . Autumn will have been postponed.

T. On the 4th, Manchester will be experiencing warm and settled
weather.

Not many problems with that forecast. At the time of issue, others
were looking at an extension to the unsettled weather and an early
autumn. T should have a very pleasant day in Manchester tomorrow and
the whole of the UK has a very nice weekend to look forward to before
the rain finally gets ino the west on sunday, heralding a change. I
did say that change would be slow in coning and so it has proven.

A good forecast. Correct 72/91 79.1%

On Sep 3, 4:48*pm, "Will Hand" wrote:

But I still come back to how useful all this is apart from an academic
exercise? Is it possible to offer a forecast service when the only time you
issue a forecast is when the confidence is high.

Isn't that something of a straw man though? AIUI he/she/it (ie
Dawlish) is simply exemplifying the unsurprising but not uninteresting
point that while 90-95% of the time no useful forecast can be made
whatsoever at 10 days distance (beyond probabiltiy estimates for the
climatologically-based means for the time of year, location etc) there
are occasions when some worthwhile forecast can be made. There's no
suggestion of trying to run a 365-day service, either for profit or
otherwise.

Rather, it's one small step towards seeing if there might be
alternative ways beyond single model extrapolations for making medium-
term forecasts. If this approach could be quantified at all (for all I
know maybe it has already) and it was possible to attach some numeric
probability values to different outcomes then maybe one might equally
find that there are periods when a 10-day forecast is so improbable as
to be impossible in practice. So you've divided up the 10-day weather
into three possible categories: moderate confidence, low confidence,
zero confidence.

And then is there another tool that might be used to add to the set.
For example, while pattern matching doesn't do too well in general at
10 days, might it be possible to identify, reproducibly, circumstances
when pattern matching could give a more confident 10-day forecast?
(I'm not arguing that it can, simply suggesting that there might be
alternative empirical approaches would could in theory be used in
concert according to their predictive relevance to a given situation
as a sort of tool-kit to improve 10-day forecasts a little further.)
And might there then be some further approach. Then slowly,
incrementally, one little step at a time, a better forecasting
strategy might be constructed.

/waffle (But science often does progress in little steps whose value
isn't easy to spot at the time.)

JGD

On Aug 24, 10:18*pm, Dawlish wrote:
Well! I can't ignore that, as my forecast criteria have been triggered
by cross-model agreement and consistency from both the gfs and the ECM
today:

** At T240, on 3rd Sept, the UK will be bathed in a warm airstream,
with a continental feed. Higher pressure will be over, or to the east
of the UK and lower pressure will be kept out in the Atlantic, or
towards Iceland, by the bocking high. This situation will have been
preceded by a *period of dry and settled weather, starting towards the
end of the last week in August and including the Bank Holiday. **

Short and simple. I don't see an early autumn at all. The unsettled
weather will continue this week, but by early September, there will be
(questionable, for sure!) talk of an "Indian summer" (whatever that
actually is!) . Autumn will have been postponed.

T. On the 4th, Manchester will be experiencing warm and settled
weather.

Do easterlies count as southerlies Paul?

Len

ah la peanut butter sandwiches!

poof! magic eh?!

1 gold star to you then you ****ing **** for stating the ****ing obvious.

your no one.
no qualification
no history

no ****ing leg to stand on

****ing charlatan copying the met office and claiming your own




03/09/2010 10:07 AM, Dawlish wrote:

A good forecast. Correct 72/91 79.1%